If User A, receives the product from Merchant B but then broadcasts the original tx for a refund,
then User A stole from the merchant the same way Credit Cards do charge backs. My understanding,
which is limited, is that the merchant has the ability to push a corresponding tx that says that he sent
the product without getting paid, and the whole tx exchange falls into a new type of escrow negotiation
contract phase. I remember hearing some weird aspects where the merchant can be paid a % in the event
that the two parties can not come to an agreement, but this is beyond my knowledge, which is limited.
but now you are taking the control away from peer-to-peer and needing a middleman controlling the keys to mediate transactions and b the only one that broadcasts a tx to bitcoin. which is no different then bank accounts or depositing funds into middlemens permissioned services
Its programmed within the LN code, so it is essentially like saying that all Ethereum
contracts and actions are also middleman controlling the keys.
LN is an automated escrow contracting system.
invalidates the previous how?
right now the only concept of invalidating it is the newest TX has a date lock then allows broadcast one block sooner.. as explained.
but then read the issues at the bottom of the last post. where that concept falls apart based on the "utopian" promotion of infitinte transactions and never closing channels
When a new purchase is created within the same LN Hub, the prior LN tx will be overwritten.
The only tx that is important is the final LN to BTC tx that settles back to the bitcoin blockchain.
All the unbroadcasted txs you are referring to, I think, only exist within the LN.
When those txs are agreed or disputed, then it executes the "master tx" that settles the account.
This is my understanding. I am not an expert nor claim to be.
will be over written? if i make a tx. i can copy it out of the LN mempool and keep it.
The only tx that is important: is the TX that gets accepted by a mining pool and confirmed in a block first.
When those txs are agreed or disputed, then it executes the "master tx" that settles the account. (the first reset tx)
i understand and have no complaints about you, and im not complaining, i actually feel you are willing to learn and think about scenarios and are trying to learn.
all im saying is LN is not a utopia of perfection some fanboys think it is. it still has flaws and people thinking about LN should recognise some limitations.
LN has got some good usecases, but should atleast admit to its limitations to reduce risk, find flaws and solve bugs. to then have something that will be useful for the usecases.
anyway ive digressed.
in the concept of LN being unlimited transactions (flaw due to locktime protection) and never needing to close(flaw of broadcasting older tx once X time passes) that it is possible to double spend. by a user having a TV and his btc back. leaving the other person with neither.
I'm not sure, but I think the LN tx system is not the same as Bitcoins.
It would have other aspects besides the proof signatures.
I think there are programmable functions that must be fully met before the tx can validate.
I think the unbroadcasted LN tx system is contingent on unfinalized agreements.
If the two parties create a new agreement or update the agreement, the prior tx doesn't execute
correctly and would send back an error that says the agreement terms are new or updated.
When all conditions are met, then they are "finalized" and then the tx can broadcast and be validated.
I don't know what the truth is, this is what my basic understanding was.
But I agree in part, that this type of system will never be as trustless and secured as Bitcoin p2p.
I think the LN Devs admit that there are issues and problems, especially with trust and don't consider it "perfect".
But it is still worth experimenting and attempting to solve these problems.
Edit: As a reminder, I was always told that this system is for small exchanges, so a TV, depending on the price,
might be too risky in the LN system. A TV worth 500 USD should stay on the bitcoin blockchain, IMO.